Signaling effect in social network and charity crowdfunding: Empirical analysis of charity crowdfunding of Sina MicroBlog in China

Abstract

With the increasing number of online charity donations, research on the influencing factors of individual donation behavior has become an important topic. Social interaction information in crowdfunding has become an essential basis for potential backers to make decisions. It provides new research space for charity crowdfunding and social capital theory. The primary purpose of this study is to explore the influence of social capital, social recommendation, and other signals on charity crowdfunding performance. We obtain 4,780 project information on the charity crowdfunding of Sina MicroBlog through data collection procedures. Our research found that both external social capital and internal capital can significantly improve the fundraising performance of crowdfunding projects. Projects with more social recommendations are more likely to obtain financial support. In the case of Medical aid crowdfunding projects, the positive promotion effect of social recommendations on project fundraising ability is enhanced. To get more effective support for crowdfunding projects, it is necessary to pay attention to the construction of social capital and the cultivation of its reputation to obtain the recognition of potential backers.

Machine Learning and Non-Investment Crowdfunding Research: A Tutorial

“…Duration is the time period during which the campaign was active. Consistent with other studies in this area we expect this variable to have a negative coefficient in a logistic regression model, i.e., this variable negatively affects the probability of success (Mollick, 2014; Bi et al, 2017; Butticè et al, 2017; Colombo et al, 2015; Courtney et al, 2017; Li et al, 2022; Lin and Boh, 2021; Shneor et al, 2021; Skirnevskiy et al, 2017).…”

“…Second, the study provides some interesting insights on the nonlinear relationships between the explanatory variables and the likelihood of success for art projects. For example, the output shows the existence of threshold and Goldilocks effects, which contradicts the constant positive or negative coefficients that the literature predicts with respect to the explanatory and control variables (e.g., Butticè et al, 2017;Colombo et al, 2015;Courtney et al, 2017;Li et al, 2022;Lin & Boh, 2021;Mollick, 2014;Shneor et al, 2021;Skirnevskiy et al, 2017;Usman et al, 2020).…”

“…Refining this, the Table shows that SCP is the most important variable with 54% main effect and 63% total effect. The importance of the goal of the campaign are demonstrated in the extant literature (e.g., Mollick, 2014; Colombo et al, 2015; Butticè et al, 2017; Bi et al, 2017; Usman et al, 2020; Lin & Boh, 2021; Li et al, 2022). As such, the table shows that the Log ( Goal ) has the second most important effects on the outcome with 15% main effect and 26% total effect.…”

“…Goal is the monetary goal of the campaign. Consistent with the extant literature, we expect this variable to have a negative coefficient in logistic regression, i.e., it negatively affects the probability of success in a logistic regression model (Mollick, 2014;Colombo et al, 2015;Butticè et al, 2017;Bi et al, 2017;Usman et al, 2020;Lin and Boh, 2021;Li et al, 2022). For scaling purposes we will use the logarithm of the goal, Log (Goal).…”

“…At the same time, academic research on crowdfunding has also grown (e.g., Deng et al, 2022; Kaartemo, 2017; Shneor and Vik, 2020; Shneor et al, 2020). The vast majority of studies on non-investment crowdfunding use parametric regression tools (e.g., Molick, 2014; Colombo et al, 2015; Butticè et al, 2017; Bi et al, 2017; Lin and Boh, 2021; Usman et al, 2020; Shneor et al, 2021; Li et al, 2022; Elitzur et al, 2023a) 1 . A number of studies use machine learning methods to study noninvestment crowdfunding (Duan et al, 2020, Peng et al, 2021; Elitzur & Solodoha, 2021; 2 Elitzur, Katz, et al, 2023; Oduro et al, 2022; Wang et al, 2021, 2022; Zhong et al, 2022) 3, 4 .…”